Trust & transparency
Recommendation
Use when the agent has spotted something and has a specific next step to propose, and a person makes the call.
Agent suggestion with confidence and clear actions.
Part of Trust & transparency. The stage where the answer has arrived and the user decides whether to believe it. These patterns show how the agent got there, what it read, and how sure it is. They matter most once people start acting on what the agent says.
Do
- State the recommendation as one sentence with the key values picked out: who, how much, by when.
- Show the other options the agent considered and how it rates them.
- After accepting, say what happens next and who does it.
Don't
- Don't recommend without saying how confident the agent is.
- Don't offer several primary actions. One accept, with alternatives as a quieter choice.
- Don't carry the action out on accept without saying so. "Queued for Dana to send" tells the user where it went.
States
- Proposed: A card with the issue, an AI suggestion badge, the recommendation, a confidence meter, the other options, and Accept.
- Alternatives hidden: The other options collapse and the secondary button reads "Alternatives".
- Accepted: The options and buttons give way to a drawn tick, "Queued for Dana to send" and a live badge.
Usage
import { RecommendationPattern } from "./halaska-kit";
<RecommendationPattern />
Built from
Agent Glyph, AI Suggestion Badge, Badge, Button, Card, Stack, Typography
Related patterns
Halaska UI · Components · Patterns · Screens · Install